Upload folder using huggingface_hub
Browse files- .ipynb_checkpoints/README-checkpoint.md +5 -5
- README.md +5 -5
.ipynb_checkpoints/README-checkpoint.md
CHANGED
|
@@ -37,17 +37,17 @@ feature_extractor = AutoFeatureExtractor.from_pretrained(model_name)
|
|
| 37 |
|
| 38 |
3. Calculate Voice Similarity
|
| 39 |
```shell
|
| 40 |
-
file_path1 = './
|
| 41 |
-
file_path2 = './
|
| 42 |
|
| 43 |
feature_extractor = Wav2Vec2FeatureExtractor.from_pretrained(model_name)
|
| 44 |
def load_and_process_audio(file_path, feature_extractor, max_length=4.0):
|
| 45 |
audio, sampling_rate = librosa.load(file_path, sr=16000)
|
| 46 |
inputs = feature_extractor(audio, sampling_rate=sampling_rate, return_tensors="pt", padding="max_length", truncation=True, max_length=int(max_length * sampling_rate))
|
| 47 |
-
return inputs.input_values
|
| 48 |
|
| 49 |
-
audio_input1 = load_and_process_audio(file_path1, feature_extractor)
|
| 50 |
-
audio_input2 = load_and_process_audio(file_path2, feature_extractor)
|
| 51 |
|
| 52 |
embedding1 = model(audio_input1).last_hidden_state.mean(dim=1)
|
| 53 |
embedding2 = model(audio_input2).last_hidden_state.mean(dim=1)
|
|
|
|
| 37 |
|
| 38 |
3. Calculate Voice Similarity
|
| 39 |
```shell
|
| 40 |
+
file_path1 = './test1.wav'
|
| 41 |
+
file_path2 = './test2.wav'
|
| 42 |
|
| 43 |
feature_extractor = Wav2Vec2FeatureExtractor.from_pretrained(model_name)
|
| 44 |
def load_and_process_audio(file_path, feature_extractor, max_length=4.0):
|
| 45 |
audio, sampling_rate = librosa.load(file_path, sr=16000)
|
| 46 |
inputs = feature_extractor(audio, sampling_rate=sampling_rate, return_tensors="pt", padding="max_length", truncation=True, max_length=int(max_length * sampling_rate))
|
| 47 |
+
return inputs.input_values
|
| 48 |
|
| 49 |
+
audio_input1 = load_and_process_audio(file_path1, feature_extractor).to(device)
|
| 50 |
+
audio_input2 = load_and_process_audio(file_path2, feature_extractor).to(device)
|
| 51 |
|
| 52 |
embedding1 = model(audio_input1).last_hidden_state.mean(dim=1)
|
| 53 |
embedding2 = model(audio_input2).last_hidden_state.mean(dim=1)
|
README.md
CHANGED
|
@@ -37,17 +37,17 @@ feature_extractor = AutoFeatureExtractor.from_pretrained(model_name)
|
|
| 37 |
|
| 38 |
3. Calculate Voice Similarity
|
| 39 |
```shell
|
| 40 |
-
file_path1 = './
|
| 41 |
-
file_path2 = './
|
| 42 |
|
| 43 |
feature_extractor = Wav2Vec2FeatureExtractor.from_pretrained(model_name)
|
| 44 |
def load_and_process_audio(file_path, feature_extractor, max_length=4.0):
|
| 45 |
audio, sampling_rate = librosa.load(file_path, sr=16000)
|
| 46 |
inputs = feature_extractor(audio, sampling_rate=sampling_rate, return_tensors="pt", padding="max_length", truncation=True, max_length=int(max_length * sampling_rate))
|
| 47 |
-
return inputs.input_values
|
| 48 |
|
| 49 |
-
audio_input1 = load_and_process_audio(file_path1, feature_extractor)
|
| 50 |
-
audio_input2 = load_and_process_audio(file_path2, feature_extractor)
|
| 51 |
|
| 52 |
embedding1 = model(audio_input1).last_hidden_state.mean(dim=1)
|
| 53 |
embedding2 = model(audio_input2).last_hidden_state.mean(dim=1)
|
|
|
|
| 37 |
|
| 38 |
3. Calculate Voice Similarity
|
| 39 |
```shell
|
| 40 |
+
file_path1 = './test1.wav'
|
| 41 |
+
file_path2 = './test2.wav'
|
| 42 |
|
| 43 |
feature_extractor = Wav2Vec2FeatureExtractor.from_pretrained(model_name)
|
| 44 |
def load_and_process_audio(file_path, feature_extractor, max_length=4.0):
|
| 45 |
audio, sampling_rate = librosa.load(file_path, sr=16000)
|
| 46 |
inputs = feature_extractor(audio, sampling_rate=sampling_rate, return_tensors="pt", padding="max_length", truncation=True, max_length=int(max_length * sampling_rate))
|
| 47 |
+
return inputs.input_values
|
| 48 |
|
| 49 |
+
audio_input1 = load_and_process_audio(file_path1, feature_extractor).to(device)
|
| 50 |
+
audio_input2 = load_and_process_audio(file_path2, feature_extractor).to(device)
|
| 51 |
|
| 52 |
embedding1 = model(audio_input1).last_hidden_state.mean(dim=1)
|
| 53 |
embedding2 = model(audio_input2).last_hidden_state.mean(dim=1)
|